Applied research

AI Project

Applied machine learning research aimed at improving decision support and widening access in healthcare.

Overview

Innovation beyond delivery platforms.

At Soin Pharmaceuticals, innovation is not limited to delivery platforms. We are also interested in how intelligent systems can help support better healthcare decisions. Our AI work reflects a broader belief that healthcare access can improve when data, clinical insight, and practical tools are brought together responsibly.

Featured project

Leveling the playing field in rural healthcare

One featured project explores how machine learning can help physicians evaluate spinal-pain conditions and better match patients to likely interventions. According to the uploaded abstract, the study involved 250 patients across multiple diagnoses, approximately 80 data points, and tracked variables such as pain scores, functional status, pain location, duration, and intervention history.

The stated goal was to create a dataset capable of helping a computer algorithm guide physicians toward therapeutic options with a higher likelihood of success.

Patients

25

Data points

~80

Variables tracked

Pain scores

Functional status

Pain location

Duration

Intervention history

Future directions

Why this matters

The abstract also highlights future directions that are highly relevant to a public- facing innovation page: narrowing the most meaningful variables, improving response quality, adding more patients, and eventually incorporating treatment cost into the model’s decision framework. Those details make the project feel practical rather than abstract, and especially relevant to settings where access to specialist expertise may be limited.

Narrowing the most meaningful variables

Improving response quality

Adding more patients

Incorporating treatment cost

Our position

Our approach to responsible AI

We view AI as a support tool, not a substitute for physician judgment. Any clinical AI system must be grounded in high-quality data, iterative validation, and careful real-world use.

Our tone stays aligned with decision support, clinical partnership, and research maturity rather than making sweeping claims about diagnosis automation.